Hey Anyone,
In a four output tube push pull amplifier is it better to use two coupling caps or four? If capacitors cause a phase shift wouldn't it be better to keep both sides on the same cap?
In a four output tube push pull amplifier is it better to use two coupling caps or four? If capacitors cause a phase shift wouldn't it be better to keep both sides on the same cap?
definitely FOUR
Each of the tubes may be at slightly different bias, therefore individual capacitors block the impinging DC independently.
GoatGuy
Each of the tubes may be at slightly different bias, therefore individual capacitors block the impinging DC independently.
GoatGuy
Assuming the correct cap values are used, phase shift is not the issue. Bias is the issue. If the valves have separate fixed voltage bias then separate caps are essential. If some other bias method is used then separate caps are not necessary, but they may still be useful as they allow DC separation of the grids. This means that a gassy valve will not get a chance to send its neighbour into thermal runaway too. A leaky cap will only kill one valve not two.
So with regard to bias the common connection to the cap allows a common DC connection between the grids?
Unless you can find a way to connect things and at the same time not connect them a common cap means a common grid connection. It doesn't allow it; it requires it.
Use one cap per tube.
DF96 has stated it 3 different ways.
Independent blocking of DC is the core.
GoatGuy
DF96 has stated it 3 different ways.
Independent blocking of DC is the core.
GoatGuy
I wasn't disagreeing with you guys at all. I was just trying to understand. In the amps now I have some stupid expensive Mundorff Silver in Oil caps. But I only have four.
Technically if the amp uses cathode bias AND there is a separate individual cathode resistor for each tube, you can use 2 caps. Note that one gassy tube can upset the bias for the other tube since the grids are tied together. You do have to divide the value of the grid leak resistor by the number of tubes sharing it.
In a Big Dumb Blonde One experiment I successfully connected EIGHT EL84's up to one 12AT7 using only 2 caps. Nothing blew up and I got over 100 watts out of it on over 400 volts. Tubes were JJ's and it was a guitar amp made out of two Simple P-P boards.
In a Big Dumb Blonde One experiment I successfully connected EIGHT EL84's up to one 12AT7 using only 2 caps. Nothing blew up and I got over 100 watts out of it on over 400 volts. Tubes were JJ's and it was a guitar amp made out of two Simple P-P boards.
I just blew a bunch of money on some CE power supply caps. So I can't afford to add four Mundorffs. The tubes were originally matched pairs. Could you use fixed bias on each side with separate cathode resistors? Or would the DC still be too far off? If not I'll just have to use plastic.
I would just use a decent commercial quality cap, then cost is not an issue. It would be funny if the cost of 'audiophile' caps meant that you had to adopt an inferior circuit arrangement!
You know, a lot of people swear by Wima MKP10. I've used them and they sound really good. For the cost they put everything to shame. But to be honest the Mundorf silver in oil caps are superior in the definition department. And I already had four of them so I just bought four more. (What would I have done with the four I have?) Its only money.
I have found that the film type capacitors are made to work at frequencies far above audio and am happy with the yellow 630V types that are sold at very reasonable prices.
They sell some white ones that are rated at 1000V that show a little bit of leakage when wired in series with a DMM across 500Vdc. The yellow types just AC couple the voltage variations in the mains supply to the meter.
If they can be used in a decoupling position in a transmitter they are up to audio and will be as good as anything else no matter what it costs.
They sell some white ones that are rated at 1000V that show a little bit of leakage when wired in series with a DMM across 500Vdc. The yellow types just AC couple the voltage variations in the mains supply to the meter.
If they can be used in a decoupling position in a transmitter they are up to audio and will be as good as anything else no matter what it costs.
- Status
- Not open for further replies.